distribution of acids between water and several immiscible solvents

3 178

RAYMOND

Vol. 54

C. ARCHIBALD

The marked rise in decomposition a t 2537 A. cannot be due to activation of mercury vapor in the absorption tube, for spectrum photographs show that with the type of high pressure arc used in this study the 2537 A. resonance line is completely reversed, only the wings and the continuum to the long wave length side, which are not absorbed by cold mercury vapor, being present. The authors wish to acknowledge their indebtedness to Dr. J. H. C. Smith for many helpful suggestions made during the course of this work.

Summary 1. The first of a series of articles on the photochemical reactions of the aliphatic aldehydes has been presented. The apparatus and method of study has been described and the experimental results given for propionaldehyde. 2. In the region investigated, A2537 A. t o A3130 A,,photochemical decomposition and polymerization are independent reactions. The quantum yield of decomposition is independent of pressure, while the quantum yield of apparent polymerization is directly proportional t o the pressure. 3. The variation of the quantum yields of both reactions with wave length has been studied. 4. The principal products of decomposition were found to be carbon monoxide and ethane. A small amount of hydrogen also was formed in decomposition. 5. Fluorescence was produced by exposure to all wave lengths down to and including A2654 A. 6. Mechanisms for both photochemical reactions have been proposed. STANFORD UNIVERSITY, CALIFORNIA [CONTRIBUTION FROY THE

ANATOMICAL LABORATORY OF

THE UNIVERSITY OF

CALIFORNIA]

DISTRIBUTION OF ACIDS BETWEEN WATER AND SEVERAL IMMISCIBLE SOLVENTS B Y RAYMOND c. ARCHIBALD RECEIVED APRIL 25, 1932

PUBLISHED AUQUST5, 1932

Part I. Monocarboxylic Acids The distribution ratios a t 25' of the six straight-chain members of the saturated monocarboxylic acid series from formic acid to caproic acid have been determined between water and the following solvents: ethyl methyl ketone, tertiary amyl alcohol, secondary butyl alcohol, normal butyl alcohol and normal amyl alcohol. Data have also been included from the literature for ratios between ethyl ether and water' and between isopropyl l

W. U. Behrens, 2.anal. Chi", 69,97 (1926).

Aug., 1932

DISTRIBUTION OF ACIDS IN SOLVENTS

3179

ether and water.2 Since the distribution ratios change slightly with the concentration of the acid, the values of the ratios for ethyl ether have been estimated for an acid concentration of about 0.001 hf in the water layer. The values for isopropyl ether and water were obtained by using weighted means from the three papers cited for an average acid concentration of both layers of about 0.05 M . These values of the distribution ratios in water and ethyl ether, and in water and isopropyl ether are included in the table and plot of the original data. The exact concentrations in the ether layers cannot be calculated because the volumes after mixing are not given. Note: The value given for butyric acid by Werkman in his first paper is undoubtedly in error.

2 4 6 Number of carbon atoms in acid. Fig. 1.-Distribution of monocarboxylic acids: 1, ethyl methyl ketone; 2, tertiary amyl alcohol; 3, secondary butyl alcohol; 4,normal butyl alcohol; 5, normal amyl alcohol; 6, ethyl ether; 7, isopropyl ether. Ratios are: moles per liter in solvent layer/moles per liter in water layer. 0

The plot, Fig. 1, shows the logarithm of the distribution ratio a t the concentration given in Table I, or in cases where two concentrations are given, the logarithm of the average distribution ratio. Experimental Industrial solvents were obtained from Shell Chemical Company, Sharples Solvents Corporation, Commercial Solvents Corporation, Carbide C. H. Werkman, Ind. Eng. Chem., Anal. Ed., 2, 302 (1930); 0. L.Osburn and C. H. Werkman, ibid., 3, 264 (1931); C. H. Werkman and 0. L. Osburn, &id., 3, 387 (1931).

3180

VOl. 54

RAYMOND C. ARCHIBALD

and Carbon Chemicals Corporation, etc. Some were first washed with 20 normal aqueous sodium hydroxide and all were of the highest purity, having been distilled in a 20-foot (6.09 m.) fractionating column described TABLE I DISTRIBUTION RATIOS OF MONOCARBOXYLIC ACIDS Concentrations in millimoles per liter Acid

Concn. Hz0 layer

Conen. solv. layer

Ratio

Formic Formic Acetic Acetic Propionic Butyric Butyric Valeric Caproic

Ethyl Methyl Ketone and Water 122.69 165.11 1I3457 443.32 1.2942 342.52 178.66 1.2050 148.26 543.51 460.45 1.1804 76.45 189.50 2.4787 32.78 166.44 5.0774 615.99 5.0129 122.88 322.36 10.211 31.57 13.43 298.89 22.260


Tertiary Amyl Alcohol and Water 114.43 130.15 1.1373 397.05 360.56 1.1012 195.31 1.4791 132.05 528.29 374.27 1.4115 198.29 3.8125 52.01 166.97 10.367 16.01 535.05 10.226 52.32 330.74 27.631 11.97 291.72 82,17 3.55


Secondary Butyl Alcohol and Water 127.81 118.30 1.0804 346.25 369.53 1.0672 192.12 151.00 1.2723 531.82 466.70 1,1395 72.21 178.46 2.4714 159.63 5.1082 31.25 498.96 5.4699 91.22 27.42 311.39 11.356 288.61 24.773 11.65


Normal Butyl Alcohol and Water 256.22 225.07 0.8486 635.54 .8097 784.93 187.57 150.72 1.2445 420.54 513.07 1.2200 57.62 186.20 3.2315 21.09 187.61 8.8957 57.66 503.50 8.7322 13.71 315.50 23.012 287.40 71.32 4.03

Av. ratio

1.320 1.193 2.4787 5.045 10.211 22.260

1.119 1.445 3.8125 10.29 27.631 82.17

1.074 1.206 2.4714 5.289 11.356

24.773

0.829 1.232 3.2315 8.814 23.01 71.3

Aug., 1932

3181


TABLE I (Concluded) Acid

Formic Formic Acetic Acetic Propionic Butyric Butyric Valeric Caproic Formic Acetic Propionic Butyric Valeric Caproic

Concn. H20 layer

Concn. solv. layer

Ratio

Normal Amyl Alcohol and Water 0.5782 143.86 83.18 .5200 489.89 254.72 .9303 164.51 153.05 488.28 458.64 .9393 186.00 3.4871 53.34 11.171 15.52 173.38 519.12 11.123 46.67 312.88 35.43 8.83 109.3 2.60 284.26 Ethyl Ether and Water 0.30 .44 1.34 4.6 17.5 75

Av. ratio

0.549

.935 3.4871 11.15 35.43 109.3

(Cf.Ref I )

Isopropyl Ether and Water 0.145 .184 .810 2.90

Formic Acetic Propionic Butyric

(Cj.Ref. 2)

el~ewhere.~Blank titrations were made on all solvents in equilibrium with distilled water a t 25'. These titrations showed no appreciable acid in either the water or any of the solvents. The acids used were high grade commercial c. P.products. The solvent, water and acid were mixed in a flask and allowed to remain, with frequent shaking, in a thermostat at 25.00 * .01" for several hours. Samples of each layer were pipetted out and titrated in duplicate against standardized 0.1 N sodium hydroxide with phenolphthalein indicator. These samples were 25 cc. in all cases except the higher concentrations of formic, acetic and butyric acids, for which 10-cc. samples were used. The values in Table I were obtained from the average of the duplicate titrations. The close check of the duplicate titrations showed that equilibrium had been attained.

Discussion This work was undertaken for its bearing on the possibility of separating sterols from vitamins by forming some ester of the sterol, such as the gluconate, which might be slightly soluble in water, and fractionating the mixture in a column which gives multiple distribution between two immiscible solvents. The construction and use of this column will be pubEvans, Cornish, Lepkovsky, Archibald and Feskov, Ind. Eng. Chem., Anal. Ed., 2,339 (1930).

3182

VOl. 54


lished later in a paper from this Laboratory dealing with the isolation of the fat-soluble vitamins. This work is also of interest in connection with the Dakin process for separating amino acids.* This process consists in extracting a water solution of the mixed amino acids with normal butyl alcohol or other solvent. The strongly ionized amino acids remain in the water layer, while the slightly ionized acids are extracted. A similar process is also being used in this Laboratory in an attempt to isolate some of the water-soluble vitamins.

Part 11. Inorganic Acids This part of the paper will consider the distribution of hydrochloric, nitric, sulfuric and perchloric acid between water and some of the nonTABLE I1 DISTRIBUTION OF INORGANIC ACIDS Concentrations in equivalents per liter, except aniline values, which are moles per 1000 g. of solution. Acid

Concn. solv. layer

Concn. Hs0 layer

Solvent: n-Butyl Alcohol 0.0105 0.0839 HC1 HCl ,0857 .4197 HCl .7%6 1.7973 HNOs .0208 0.0856 HNOs ,1455 ,3657 .9841 1.5481 HNOa Has04 .0056 0.0937 &SO4 .0387 .4745

H2SOd

.3412

Acid

'

2.3416

HClOd .0353 0.0774 HClO4 * 2088 ,3224 1.1271 1,4065 HCIO, Solvent : n-Amyl Alcohol HCI HCl HCl HNOJ HNOt HNOi

0,0026 .0279 .4213 .Go71 .OS29 .7021 .0018

H & %

HSOc

.0059 .0787

0.0929 .4836 2.1964 0.0974 .4794 1.8708 0.1007 .5052 2.4714

Solvent: Twt.-amyl Alcohol HC1 0.0109 0.0956 HCI .lolo 0.4385 HCI ,9616 1.7882 HzSO4 .0056 0.0984 Hnsoc .W64 .4923 HaSOc .4869 2.2337 4

H.Dakin, Biochem. J., 12,290(1918).

Concn. solv. layer

Concn. HzO layer

Solvent: Methyl Ethyl Ketone HC1 0.0024 0.0695 HC1 .0117 ,3438 HCl .1497 1.3749 HNOI .Go88 0.0647 HNOa .0935 .2871 HNO, Miscible Solvent: Ethyl Ether HNOa 0.0011 0.0847 HNOi .0165 ,4326 HNO, ,4263 1.9071 Solvent: Aniline (Cf.Ref. 5) 0.1110 HC1 0.00599 HC1 .01939 .2039 HCl .04190 .2892 HC1 .04886 ,3117 HC1 ,05758 ,3361 HCl .09151 .4240 HCl .1134 ,4594 HC1 .1816 .5514 HC1 .3371 .7126 HCI .a558 1.0247 HCl 1.1299 1.1622 HC1 1.1876 1.1832 HCI 1.446 1.3320 HCl I.578 1.578

Aug., 1932


3183

aqueous solvents used in Part I. Distributions of these acids were determined in three concentrations, the average concentration for both layers being about 0.05, 0.25 and 1.25 N . The layers were approximately equal in volume. The experimental method is identical with that used for the monocarboxylic acids; 25-cc. samples were used in the two lower concentrations and 10-cc. samples in the highest. The data on hydrochloric acid in aniline are due to Sidgwick, Pickford and W i l ~ d o n . ~ 1

-3 -2

-1 0 1 Loglo concn. in water layer. Fig. 2.--Distribution of inorganic acids. Concentrations are in equivalents per liter, except auiline values where hydrochloric acid is in moles per 1000 g. of solution.

Table I1 shows the acid concentration in the water layer, and in the solvent layer in equilibrium with it. All measurements are a t 25.00 f .Ole. In Fig. 2 the logarithm of the acid concentration in the aqueous layer is plotted against the logarithm of the acid concentration in the solvent layer. 6 N. V. Sidgwick, P. Pickford and B. H. Wilsdon, J. Chem. Soc., 99, 1122 (1911).

3184


VOl. 54

Discussion The tendency is for the curves t o approach the line denoting unit distribution ratio as the acid concentration is increased. They reach this line a t the point where complete miscibility between the two layers occurs, which in the case of nitric acid in ethyl methyl ketone and water is within the range of concentrations studied. The curves may cross this line before the point of complete miscibility, returning t o it later, as in the case of* aniline and hydrochloric acid. The curve for hydrochloric acid in aniline and water has a slope of about 2 over most of its length. That is, the concentration of the hydrochloric acid in the aniline layer varies approximately as the square of the acid concentration in the water layer. From this fact i t may be assumed that the hydrochloric acid is practically un-ionized in the aniline layer. This condition is approximated also in the curve representing nitric acid in ethyl ether and water. The other curves, however, apparently show more tendency for the acids t o be ionized in the non-aqueous layers. Distribution measurements were also attempted for hydrochloric acid between ethyl ether and water. However, due to the extreme insolubility of hydrochloric acid in ether a t all the concentrations studied, it was impossible to make any accurate measurements without a much more refined procedure. I n any one solvent the order of the curves from the line of unit distribution ratio is HC104, HNOI, HC1, HzS04, which follows the Hofmeister series for the precipitation of proteins, the sulfate ion being the most efficient. The series is also the same as that found by Randall and Failey6 for the salting-out effect, rather than the strength of acids series’ which gives the order HN03, HC1, HzS04,HC101, where perchloric is the strongest acid. An attempt was made to measure the distribution of hydrochloric acid between propylene oxide and water, but due to the rapid reaction of propylene oxide with hydrochloric acids even in cold dilute aqueous solution, these measurements were impossible. Propylene oxide is more reactive than propylene glycol and consequently much more reactive than ethers such as ethyl ether. This extreme reactivity is probably due t o the fact that the oxygen atom is one member of a three-membered ring, and this makes propylene oxide more suitable as a reagent than a solvent. The purpose of this research was to determine quantitatively which acid would be most suitable for acidifying a water solution, one component of

* Randall and Failey, Chem. Rev., 4, 285 (1927); ibid., 4, 271 (1927); ibid., 4, 291 (1927). 7 Hantzsch and Voigt, Ber., 62B, 975-984 (1929); Hantzsch and Langbein, Z . anorg. allgem. Chem., 204, 193 (1932); Hall, The Nucleus, 6, 87-88, 91 (1929). 8 Henry, Bull. acad. roy. Belg., p. 397 (1903); Michael, Ber., 39, 2786 (1906); Smith, 2. physik. Chem., 93, 59 (1919); Markownikoff, Compt. rend., 81, 799 (1875); Nef, Ann. 335,205 (1904).

Aug., 1932

KINETICS OF GAS EXPLOSIONS

3185

which was to be extracted out with one of the solvents mentioned above. Since i t was necessary to extract with very large amounts of solvent, many times it was imperative that the acid be very insoluble in the solvent in order that the solution remain acid until the end of the extraction process. Acknowledgment.-The author acknowledges his indebtedness to Dr. Herbert M. Evans and to Dr. R. E. Cornish of this Department for their valuable assistance and advice. Summary Distribution ratios of six fatty acids and four inorganic acids were measured between water and several solvents which are not miscible with water. A short discussion of the results is given. BERKELEY,CALIFORNIA

[CONTRIBUTION FROM

THE

PITTSBURGH EXPERIMENT STATION, U. s. BUREAU OF MINES]

THE KINETICS OF GAS EXPLOSIONS. 111. THE INFLUENCE OF HYDROGEN ON THE THERMAL DECOMPOSITION OF OZONE SENSITIZED BY BROMINE VAPOR, AND THE DETERMINATION OF THE EXPLOSION TEMPERATURE' BY

w.FEITKNECHT~ AND BERNARD LEWIS3

RECEIVED APRIL2.5,1932

PUBLISHED AUGUST5, 1932

In a previous paper4 it was shown that inert gases exert a small influence on the rate of decomposition of ozone in the presence of bromine as long as the ozone pressure is low. -4t higher pressures the rate is slowed down and the explosion pressure limit of ozone is increased considerably. The efficiency of inert gases in increasing the explosion limit decreases with increasing molecular weight. The influence of hydrogen was not included in the previous investigations. It seemed of special interest, as hydrogen increases considerably the photochemical rate of decomposition of ozone with the simultaneous formation of water.6 Belton, Griffith and McKeown6 found a similar inffuence of hydrogen on the thermal decomposition of ozone, whereas Schumacher' states that hydrogen has only a slight effect on the thermal Published by permission of the Director, U. S. Bureau of Mines. (Not subject to copyright } a Formerly Foreign Exchange Fellow, University of Pittsburgh. Physical Chemist, U. S. Bureau of Mines, Pittsburgh Experiment Station, Pittsburgh, Pa. Part I, THISJOURNAL, 53,2910 (1931). Part 11, &id., 54, 1784 (1932). Weigert and Bohm, Z. physik. Chem., 90, 189 (1915); Griffith and Shutt, J . Chem. Soc., 123,2752 (1923). Belton, Griffith and McKeown, ibid., 126,3153 (1926). 'I Schumacher, THIS JOURNAL, 52,2388 (1930).

distribution of acids between water and several immiscible solvents

Recommend Documents